RoCE​/Infiniband Network Engineer

At Kyndryl, we run and reimagine the mission‑critical technology systems that drive advantage for the world’s leading businesses. We are at the heart of progress; with proven expertise and a continuous flow of AI‑powered insight, enabling smarter decisions, faster innovation, and a lasting competitive edge. For our people—Kyndryls—that means doing purposeful work that powers human progress. Join us and experience a flexible, supportive environment where your well‑being is prioritized and your potential can thrive.

Who We Are

At Kyndryl, we run and reimagine the mission‑critical technology systems that drive advantage for the world’s leading businesses. We are at the heart of progress; with proven expertise and a continuous flow of Artificial Intelligence (AI)-powered insight, enabling smarter decisions, faster innovation, and a lasting competitive edge. For our people - Kyndryls - that means doing purposeful work that powers human progress.

Join us and experience a flexible, supportive environment where your well‑being is prioritized and your potential can thrive.

As an RoCE / Infiniband Network Engineer, you will provide hands‑on troubleshooting and change execution for high‑speed Graphics Processing Unit (GPU) fabrics. You will work with Remote Direct Memory Access over Converged Ethernet (RoCE) and/or Infini Band environments, restoring service for common link, switch, Host Channel Adapter (HCA), Network Interface Card (NIC), Remote Direct Memory Access (RDMA), congestion and configuration issues.

You will own service restoration and elevate architectural or highly complex fabric conditions to the appropriate Subject Matter Expert (SME).

• Troubleshoot switch, HCA/NIC, port and optic issues using the Command-Line Interface (CLI), telemetry and fabric-management tools.
• Analyze link errors, drops, flaps, speed and Forward Error Correction (FEC) conditions, congestion counters and RDMA reachability.
• Operate Infini Band/Unified Fabric Manager (UFM) or NVIDIA Spectrum‑X/RoCE management tooling within approved access boundaries.
• For RoCE environments, troubleshoot Data Center Bridging (DCB), Priority Flow Control (PFC), Explicit Congestion Notification (ECN), Quality of Service (QoS) configuration and common lossless‑network issues.
• For Infini Band environments, troubleshoot link, subnet and fabric conditions, HCA state and common subnet‑management issues.
• Participate in Border Gateway Protocol (BGP), Ethernet Virtual Private Network (EVPN), leaf‑spine and fabric change implementation as defined by site architecture.
• Correlate network and fabric symptoms with multi‑node GPU/NVIDIA Collective Communications Library (NCCL) workload degradation and engage compute/software teams as needed.
• Perform approved firmware and Operating System (OS) changes, cable/optic replacement and post-change validation.
• Own Service‑Level Agreement (SLA) and ticket progression, contribute to Root Cause Analysis (RCA), mentor Level 1 (L1) engineers and improve operational runbooks.

This role gives you the opportunity to deepen your expertise in high‑speed networking for GPU, AI and High‑Performance Computing (HPC) environments while expanding your capabilities across RDMA, RoCE, Infini Band, fabric telemetry and automation. As your experience grows, you can progress toward senior network engineering, fabric specialist, architecture or technical leadership roles, supported by continuous learning and hands‑on experience in mission‑critical environments.

You’re good at what you do and possess the required experience to prove it. However, equally as important, you have a growth mindset and are keen to drive your own personal and professional development. You are customer‑focused,…